Five Things You Need to Know About Your Customers to Make Your Startup a Success

Artem Mariychin
Zodiac Metrics
Published in
8 min readNov 27, 2017

After I graduated from the Wharton School and entered the finance world, I began my career as an investor at firms including Goldman Sachs, 3G Capital, and Highbridge Capital. When I met with management teams it continually surprised me that the largest, most successful organizations in the world often had only the most rudimentary understanding of their own customers. Companies couldn’t tell me how much a customer would spend over their lifetime, nor how much they were spending to acquire customers across channels. Without that information, it was impossible for them to tell me the ROI of their marketing programs, to justify that their investments were creating value for shareholders, and, more generally, what was happening operationally in the business.

After seeing this issue time and time again, I got together with Wharton Professor Peter Fader and alumnus Dan McCarthy (now a professor at Emory) to figure out how we could provide businesses with a better understanding of their customers. Pete and Dan saw the problem from a different perspective. For over 20 years, they had been developing predictive models for customer behavior and working with a wide variety of companies to help them become customer-centric. We wanted to make this easy for organizations of any size: that’s why we founded Zodiac.

When we started the company in 2015, our goal was simple: provide the most accurate customer lifetime value (CLV) predictions to businesses so they could make the right strategic and tactical decisions. Why did we think this was so important?

Knowing what each customer is going to be worth to your company — the cash flow they will generate over their entire relationship with your firm– is the most important information to your business.

With that knowledge, you can focus on acquiring long term, loyal customers, offering them the right services and products, and retaining them more effectively. This approach can help you build a sustainable strategic advantage, a brand, and a long-lasting business.

This idea seems simple, but many companies get it very wrong.

At Zodiac, we analyze a tremendous amount of customer behavioral data across all types of firms: up and coming e-commerce companies, Fortune 500 omnichannel retailers, pharmaceuticals, telecommunications firms, and more. We have uncovered some nearly universal truths about customers that may surprise not just founders, but marketers, finance professionals, and investors — because this information is relevant to all of them as well. Unfortunately, we’ve seen a number of examples where companies have failed to pay attention and suffered disastrous results, such as layoffs, dilutive capital raises, fire sales, and bankruptcy.

To that end, below are five takeaways that would help any company, but especially growth-stage B2C firms.

1. Focus on your top customers and let the chips fall where they may for all the rest: Your customers are not all the same. You intuitively know this, and it may seem obvious, but there is tremendous variation in the value of your customers. You may think that customer value is “normally distributed,” where there is an “average” customer and an equal number that are better or worse, but that couldn’t be farther from the truth. Customer value is power law distributed, where most customers are very low-value and a small fraction are responsible for most of the value of a brand (the pareto principle or 80–20 rule). A high-value customer may be worth thirty or even one hundred times as much as a low-value customer. A lot of companies spend a disproportionate amount of resources catering to low-value but high-effort customers. This is often a negative ROI activity when you consider all the direct marketing costs, labor hours, etc. involved. It is in your best interest to redirect those resources toward making your high-value customers love your brand even more. Turn them into brand promoters (high NPS); ensure the brand and creative reflect their preferences; and make sure the product continues to meet their needs.

2. Let one-time shoppers go: Here’s a fact that will probably horrify most startup executives: For pretty much every company that we have seen, even those having what are considered “incredibly loyal” customers, 50–80% of customers transact once and never come back.

You may have already looked at repeat rates by cohort and seen a similar pattern. You might have even established a quarterly target to increase those rates. For most marketers, their first instinct will be to offer discounts. Don’t do it! You’ll erode the margin structure of your business just for a small upward blip in purchase activity. You’ll also train your customer base to become discount-sensitive, which will negatively affect long-term brand value.

Instead of trying to turn every customer into a loyal, repeat buyer (a noble goal, for sure, but not truly feasible), focus on understanding who your high-value, loyal customers already are. Not only will you generate a higher incremental ROI in the near-term, but you’ll learn what behaviors made them higher value in the first place. This will allow you to change the overall experience for your customers and create a journey that makes it easier for them to become high-value.

For example, let’s say you discover that your high-value customers tend to make that second repeat order within the first 60 days. While it may not be profitable to run display retargeting 100% of the time for all customers, you can do so for this very specific subset. You may also follow up with a targeted email 45 days after the initial purchase recommending additional products. Choosing these right moments to engage with customers, during “value inflection points” as we call them at Zodiac, can create tremendous ROI for your business.

3. Obsess over the ROI of your acquisition programs on a granular basis: If you are a typical e-commerce company, you probably need two transactions to break even on a customer, and your paid customer acquisition cost is in the range of $50-$100. Let’s say you have an average order value between $50–150 and a merchandise margin of around 50%. So, you’re making $25–75 per order. Given the point above about one-time purchasers, this probably makes you a bit nauseous.

You must check how much you are spending to acquire customers on average. If you’re losing money on the first transaction, and most of your customers will only transact once, you are losing money on most of the customers you acquire. The only reason your acquisition program has a positive ROI is because of the few high-value customers (those in the right tail of the distribution in point 1) that you happen to pick up. If this is true for you, it’s imperative to focus on those high-value customers and how they differ from the low-value set. Then, consider what you can do to shift your branding, creative, and targeting to attract more of them.

4. Keep in mind that the future looks nothing like the past: When it comes to customer acquisition, you are likely relying on historical cohorts to influence your decision on where to allocate your acquisition budget next period. However, despite your best efforts, the customers you acquire will change quickly. At the start of your business many of your customers may have come from channels like word-of-mouth and organic search. Once you expand your acquisition efforts and incorporate paid digital channels like Facebook and Google PPC, your customer base will begin to look very different.

As you increase your acquisition budget (especially as you raise capital), keep in mind the shifting mix between channels and the resulting change in the lifetime value of customers being acquired across them. Not only is your cost per acquisition going to increase dramatically as you pick up an increasing proportion of customers through paid channels, but their customer lifetime value is likely to differ significantly as well. Do not base budgets on your original customer cohorts, as that is an almost certain way to miss projections and burn capital. To find examples of this mistake, you can look to companies like Fab.com, Blue Apron, and Zulilly.

For this company, customers acquired through organic, free channels like SEO/organic search were worth significantly more compared to customers acquired through paid channels like AdWords.

5. The customers you acquire tend to get worse over time: This is the hardest truth of all, but you may have already acquired your best customers. Logically, this makes sense: you started out selling through friends and family, word-of-mouth, and referrals from highly satisfied initial customers, all of whom identified with your brand core. But now you’re pushing further into the edges of your target market. These customers are less loyal to you; they won’t identify with the brand proposition in the same way; and they are likely to have a lower lifetime value. This downward trend will probably continue as your customer base grows.

Below are a few examples of real data depicting this phenomenon. In each, the x-axis represents the period of acquisition (when the customer first made their transaction) and the y-axis the average lifetime value of customers in that quarter.

Note: Zodiac defines lifetime value as the present value of all spend from the first transaction until the customer churns from the brand, including transactions predicted to happen in the future. As a result, it is an apples-to-apples comparison of older and younger tenured customers. Note: all of the CLV numbers below have been randomized for confidentiality purposes, but the trends have been retained.

It’s important to keep this fact in mind as you consider budgeting, operational planning, headcount growth, etc., or it will be easy to miss your internal sales forecasts, over-hire, and put the company in a vulnerable position competitively.

For this online retailer, average CLV decreased from almost $600 in 2010 to around $200 today.
For this online travel company, average CLV decreased from ~$400 for customers acquired in 2007 to around ~$250 today. At the same time, the competition from OTA’s and their sophistication has increased, causing the acquisition cost to increase.
For this online retailer, average CLV decreased from over $2k in 2014 to less than $1k in 2016.
For this omnichannel retailer, average CLV decreased from ~$1100 in 2010 to ~$650 today.

The Silver Lining

These five facts may seem negative, but they don’t need to be. Many companies, not just startups, fail because they don’t understand how to analyze customer behavior and value. However, if you can master these concepts now, you’ll be able to build a structural, sustainable advantage for your business by building a moat around your best customers.

A company that has customer lifetime value at the core of its strategy is much more likely to understand these effects and base an effective strategy on them. This involves learning as much as possible about your highest value customers, which channels they shop in, what brand values they identify with, and which products they care about — then working to align with those traits. Companies that implement these customer-centric strategies grow more sustainably, enjoy higher profits, and build brands that truly delight their loyal customers.

--

--

Artem Mariychin
Zodiac Metrics

Co-founder and CEO of Zodiac. Passionate about the use of data to improve decision-making. Former investor at Goldman Sachs, 3G Capital, and Highbridge Capital.